Translational control is a major mode of gene regulation in eukaryotes that is often perturbed in human diseases. Despite extensive mechanistic studies of translation, it has not been possible to monitor in vivo most steps of translation initiation, which is the primary point of translational control. As a result, how endogenous translating messenger ribonucleoprotein particles (mRNPs) are formed and regulated in the competitive cellular environment is not completely understood. The goal of this application is to develop and apply in vivo methods to study the complex process of translation initiation. By revealing fundamental mechanisms of translational control, the proposed studies will provide new insights into how aberrant translation contributes to diseases, particularly cancer. In addition, the experimental systems established in this application can be widely adapted to studying translation initiation in different cellular and disease states. The methods developed in this application are based on the premise that the translating mRNP contains many mRNA-binding proteins, so individual steps of initiation can be studied by identifying and characterizing the underlying mRNA-protein interactions. This novel approach to the translating mRNP leverages in vivo crosslinking to capture RNA- protein interactions as they occur in the cell, and high-throughput sequencing to identify these interactions across the transcriptome. The experiments proposed in this application will investigate three aspects of the translating mRNP in yeast and mammalian cells. The first specific aim focuses on how the closed-loop structure is formed. Translating mRNPs adopt a circular conformation that enhances translation and mRNA stability, but how this closed-loop structure initially forms is not understood. High-throughput methods will be developed to detect endogenous closed-loop structures in yeast and mouse embryonic stem cells (mESCs). These methods will be combined with reverse genetics, chemical inhibitors, and reporter assays to determine how known protein-protein interactions and the process of translation itself contribute to closed-loop formation. These studies will provide a technical and conceptual framework for molecular analysis of the closed loop in vivo. The second major aim is to identify what determines the efficiency with which mRNAs are translated, which varies widely among endogenous genes for largely unknown reasons. The studies in this application will test the hypothesis that translational efficiency (TE) is largely explained by recruitment of the eIF4F cap- binding complex, which is the first step of initiation. Transcriptome-wide measurements of eIF4F binding and TE, combined with biochemical analysis of eIF4F-mRNA interactions, will be used to evaluate this hypothesis. Comparing results obtained in yeast, mESCs, and embryoid bodies (derived from mESCs) will reveal the extent to which principles of translational control differ between organisms and between cell types. The third aim of this project is to study the scanning 40S ribosome, which has not previously been observed or directly assayed. A method will be developed to identify the positions of isolated 40S ribosomes across the mESC transcriptome. Examining how scanning ribosomes are affected by chemical inhibition of the eIF4A helicase will provide new insights into the molecular details of scanning and the role of eIF4A in this process. Collectively, these aims will improve our knowledge of cellular translation mechanisms and our understanding of how these mechanisms go awry in disease.